Head in the Cloud
Apr 15th, 2008 by sharps
A couple of related articles hit my (nicely organized) feed reader this morning. First, Tim Bray paints a bleak picture of Google’s Developer Platform - some concerns I shared earlier. Meanwhile CogHead are in the news with their Open Definition model - an attempt to build an open and collaborative environment for CogHead customers. Interesting but CogHead is still fundamentally a sharecropping model - once you’ve committed to CogHead - your stuck with it. I’m not saying that’s all bad. It’s a tradeoff - leverage CogHead’s f5g awesome platform but box yourself in. Sound familiar ?
It’s interesting that we’re seeing (at least) three models emerge :
- Amazon’s utility computing approach (they’re fundamentally a bandwidth, MIPS and storage utility) - you choose what platform you want to write to (Linux, Java, etc.);
- the proprietary SaaS platform approach (SalesForce, CogHead - sorry guys). You trade convenience, agility for complete dependence on the vendor.
- a hybrid - Google App Engine, Facebook. I don’t think these hybrids have anywhere near the ability to lock you in as much as the true SaaS platforms. For example - Google is based on Python, Django - if you’re smart you won’t get locked in. Many facebook apps. also support other ‘distributions’ - so it isn’t really a sharecropping model.
As I see it - the only thing that might lock you in to Google and Facebook’s model is convenient access to their huge customer base - that’s what they’re selling. Take it or leave it.
Rich, I was fortunate to be one of the original guys at Red Hat where we pioneered the open source business model. We demonstrated that being “open” accelerates adoption and innovation and that translates into business benefits. When I came to Coghead, I wanted to infuse our model with the benefits of “open” while also giving people the maximum in convenience and agility. At times, these things might seem in conflict, but by bringing some innovative thinking to bear, I think we can be both open and convenient.
Our Open Definition model is but one example. We have also published our APIs, and we plan to open our widget interface and our integration framework to allow third parties complete access. We also make it really easy for someone to export their data and we auto-generate a web services interface for every application that is created on Coghead.
Paul McNamara
CEO, Coghead
Thanks for the comment Paul. Just to be clear - I think CogHead rocks - I played a little on the original beta prog. I’d really like to think that what you’re pioneering is the future of app. development for the web - I’ve been looking for someone to give people a reason to stop churning out code. Coghead comes close IMO.
Being Open has many meanings - in this context I was really focussing on the price of entry (and exit). No matter how open your platform is - the probability of sucessfuly moving a CogHead application to another environment is, right now about zero (correct me if I’m wrong). As long as CogHead provides value above the competition - few will care; but history has taught us to be catious.